If supported, this three-year project will develop new theoretical and empirical models to better understand the timing, magnitude, and consequences of intergenerational transfers. Our theoretical work will develop models that can better match empirical facts about the timing of transfers within the family, and can help explain the prevalence of """"""""tied"""""""" educational and housing transfers. We also develop potentially innovative ways to assess circumstances under which cooperative models better match behavior than non-cooperative models. Better underlying behavioral models of intergenerational transfers will enhance researchers' and policy-makers' understanding of intergenerational linkages within families, human capital formation of young Americans, and the way public policies and families interact to affect behavior. Our proposed empirical work focuses on 1) the role of tied transfers in understanding college attendance and graduation, 2) empirical tests of the implications of cooperative models of transfers, 3) the effects of retirement on inter-vivo transfers, 4) the effects of estate and gift tax changes on inter-vivo transfers, and 5) the joint estimation of inter-vivo transfers of time and money and related projects exploring the degree to which transfers are compensatory. The empirical projects test specific implications of new behavioral models of intergenerational transfers, and hence can demonstrate (or falsify) their ability to better understand behavior. Equally importantly, we address practical policy questions about the degree to which retirement affects transfers, about the effects of estate and gift taxation on inter-vivos transfers, and about the degree to which families efficiently allocate resources across family members. Our empirical work will be based on complementary transfer questions from the Health and Retirement Study (and AHEAD), the Wisconsin Longitudinal Survey (WLS), and the Surveys of Consumer Finances. The co-investigators are helping to design new transfers information in the 2003 wave of the WLS and are ideally situated to analyze these data.
